Assessment of social alienation: Psychometric properties of the SACS-R

The psychometric properties of a revision of the Social Alienation from Classmates Scale for college students was examined. A number of constructs, and related measures, have been developed to address estrangement. Consistent with prior research on other measures of estrangement, SACS-R scores were expected to be correlated with scores on tests of depression, stress, introversion, attractiveness, and family functioning. Social Desirability was controlled through partial correlations, and a peer sociometric was also employed. Results support the reliability and validity of SACS-R scores from a sample of college men.

A number of constructs, and related measures, have been developed to address the broad-band concept of social estrangement (Robinson, Shaver, & Wrightsman, 1991). Loneliness, one such construct within the estrangement area, has been empirically related to depression Wilson & Lavelle, 1990; Brage, Meredith, & Woodward,1993), stress symptoms (DeBerard & Kleinlnecht,1995), introversion (Hojat, Erdman, Robeson, & Damjanov, 1992), negative ratings of physical attractiveness (Veto, 1992), and lower family cohesion (Carver & Jones, 1992). Social Alienation is another construct under the estrangement umbrella, closely related to loneliness. Seidel and Vaughn (1991), for example, developed the Social Alienation from Classmates Scale for secondary school students as a measure of "the individual's perception of being rejected or perceived unfavorably by others" (p. 152).

Self report measures of negatively perceived characteristics such as social alienation are subject to social desirability response bias. Social desirability responding has been negatively related to scores on self report tests of loneliness (Borys & Perlman, 1985), depression (Tanaka-Matsumi & Kameoka, 1986), and stress (Koman, 1991), and positively related to measures of extraversion (Rahim, 1984). Correlations among these self report measures may, then, be inflated by shared response bias. In the current study, social-desirability bias will be measured and statistically controlled through partial correlations.

Sociometric measures, though subject to measurement biases (Epkins, 1994; Walker, Ilardi, McMahon, & Fennell, 1996), tend not to elicit social desirability bias. Thus, a correlation between an estrangement measure, such as the SACS-R, and a sociometric is not likely to be inflated by shared measurement factors. In sociometric measurement, a large number of peers are asked to report about a subject's characteristics. Research suggests that students whose sociometric ratings are lower (unpopular) tend to report more loneliness (Asher, Hymel, & Renshaw, 1984; Upmanyu, Upmanyu, & Dhingra, 1988). Persons with lower sociometric ratings are expected to obtain higher scores on the SACS-R.

Seidel and Vaughn (1991) developed the Social Alienation from Classmates Scale for use with secondary school students. They found, for example, that dropouts with learning disabilities reported more social alienation than did persisting students with learning disabilities. The revised scale for college students has been found to predict college attrition in a longitudinal study (Daugherty & Lane, 1999). Social alienation from classmates may thus predict important outcomes. The purpose of the current study is to examine the reliability and validity of scores from the revised scale administered to college men.

METHOD

PARTICIPANT

821 men at a single-gender college participated in the study. Participants, who provided informed consent, volunteered to complete psychological measures. The age range of the sample was 18 to 24 (M=20.6, SD=1.2) years.

APPARATUS

A revision of the Social Alienation from Classmates Scale (Seidel & Vaughn, 1991), a 26 item self-report inventory, was created for use with college students. Word changes were made to reflect developmental level, but item content and format remained identical to the original scale. Each item presents a statement about peer relationships, and subjects are asked to indicate how true the item is for them on a scale of 1 (never) to 4 (very). The item average is taken as the SACS-R total score, with higher scores indicating greater alienation.

Several self report tests were included in the protocol. The Marlowe-Crowne Social Desirability Scale (MCSD; Crowne & Marlowe, 1960) and the Beck Depression Inventory (BDI; Beck, 1967) were completed. The Eysenck Personality Questionnaire (EPQ; Eysenck & Eysenck, 1975) was administered, yielding measures of extraversion, neuroticism, and psychoticism. Extraversion (sociability) is expected to be negatively related to SACS-R scores, while neuroticism (instability under stress) and psychoticism (antisocial tendencies) are expected to be positively related to SACS-R scores.

The Family Adaptability and Cohesion Evaluation Scales, Third Edition (FACES III; Olson, Portner, & Lavee, 1985) provided an index of family cohesion - that is, emotional closeness and bonding among family members. Both positive (eustress) and negative (distress) emotional reactions to self disclosure were measured using Lazarus' Stress Questionnaire (Folkman & Lazarus, 1985).